Adaptive decoupling planning method for the product crowdsourcing design tasks based on knowledge reuse

Abstract

Crowdsourcing platforms are an effective way for enterprises to improve innovation abilities, reduce design costs and shorten design cycles during the boom of mass personalized customization. However, the absence of intelligent decoupling planning methods is a decisive reason for the poor intelligent decomposition of design tasks on crowdsourcing platforms. Additionally, their accumulated vast design experience and data, which should have been sufficiently reutilized to bridge that gap, have been neglected. Therefore, an adaptive intelligent decoupling planning method for relatively complex product crowdsourcing design tasks is proposed in this study, simultaneously considering knowledge reuse and bilateral feature matching. First, the requirement-knowledge-resource (RKR) characterization model is established according to the extraction of historical design experience to lay a foundation for the knowledge reuse. Second, based on the knowledge reuse, a hierarchical decoupling solution framework merged with the improved HITS-NetClus hybrid algorithm is built to perform retrieval reuse of knowledge and decoupling planning of tasks. Third, the resource competitiveness feedback strategy is taken to refine the decoupling planning scheme to meet resource competitive conditions. Finally, we demonstrate the intelligence of incorporating the hierarchical decoupling solution framework to achieve automatic decomposition using a real crowdsourcing design task for the personalized sewing machine. Our method also improves the resource adaptation capability of task decomposition schemes in practice.